Private Information, Liquidity, and Financial Crises

University Of Chicago, Chicago IL

Investigators

Abstract

Abstract Title: Private Information, Liquidity, and Financial Crises Proposal No: SES - 1326068 Principal Investigator: Shimer, Robert This research will develop and explore theoretical models of private information in financial markets. In many situations, the seller of a financial asset, such as a mortgage-backed security, may have information about the asset's quality that is unavailable to potential buyers. For example, a mortgage securitizer may be affiliated with the company that originated the mortgages that underlie the security. In this case, an investor who is considering purchasing a mortgage-backed security from the securitizer will be aware that the seller may have soft information about the homeowners' characteristics. Since this soft information affects the seller's willingness to sell, it will also affect the investor's willingness to purchase the security and hence the functioning of the mortgage market. The starting point for this research is an analysis of an environment in which there are gains from trade in financial assets. For example, a mortgage originator needs to sell some of its mortgage portfolio in order to raise money for new loans; or a mortgage securitizer needs to sell some of its existing securities in order to raise money to purchase to new mortgage pools. A buyer would like to purchase from a seller like this, with a strong intrinsic reason to sell. At the same time, sellers may have private information about the quality of the assets they own. A buyer would prefer not to purchase from a seller who is selling because he believes he has a low quality asset. Buyers are unable to tell whether a seller is selling an asset because the seller needs to raise funds or because the seller believes the asset has a low quality. Preliminary research establishes that the seller's asking price gives the buyer some of the seller's private information both about the asset's characteristics and about the seller's characteristics. The key insight is that in markets with a shortage of buyers, sellers can use their asking price as a costly signal. Sellers know that the shortage of buyers is more acute, and so a sale is less likely, at a higher price. Sellers who value holding onto their asset more, either because they perceive that the asset has a higher quality or because they do not have a strong intrinsic reason to sell the asset, are more willing to set a higher price at the risk of failing to sell their asset. Buyers know this and are willing to pay a high price to the extent that they expect to be compensated with a high quality asset. In equilibrium, the extent of the private information problem determines the shortfall in the number of buyers and hence the illiquidity of the market. More precisely, the equilibrium depends on the joint distribution of two types of private information: sellers' perception of asset quality and sellers' intrinsic reason for sale. The theoretical environment generates a rich set of empirical predictions. For example, identical assets may trade at different prices, with the price dependent on the owner's characteristics. On the other hand, heterogeneous assets may trade at the same price, so the price only imperfectly reveals an asset's quality. Some investors may simultaneously buy and sell assets in an effort to take advantage of their private information. The theoretical model suggests two notions of a crisis that leads to a breakdown in financial intermediation. The first involves a change in fundamentals. Sellers may get bad information about the quality of the assets in their portfolio. For example, a decline in house prices may increase the default risk for mortgage-backed securities that previously appeared to be safe. Alternatively, sellers may realize fewer investment opportunities, which lowers the gains from trade. For example, a decline in house prices may reduce the demand for new mortgages, moderating sellers' need to sell assets for cash. Both of these shifts imply that it is more likely that a seller is selling an asset because the seller has bad information about the asset, worsening the adverse selection problem and potentially leading to a collapse in trade. The second notion of a crisis involves a shift in behavior without a change in fundamentals. Preliminary research establishes conditions under which multiple equilibria can exist, with each indexed by the number of buyers in the market. Sellers pricing incentives are always determined by their desire to signal the quality of their asset. In an equilibrium with fewer buyers, sellers are perversely induced to set higher prices for their assets, which in turn drives out the buyers. This type of crisis may occur because of contagion which carries a crisis from one market (mortgage-backed securities) to another (securities backed by auto loans). The solid theoretical foundation that this research project will develop should be important for understanding and quantifying the importance of private information in financial markets. This in turn will be useful for evaluating the potential impact of politically feasible policy responses to a crisis and for formulating an optimal policy response